Instructions for use

Title Loss of Function of the E1-Like-b Gene Associates With Early Flowering Under Long-Day Conditions in Soybean

Author(s) Zhu, Jianghui; Takeshima, Ryoma; Harigai, Kohei; Xu, Meilan; Kong, Fanjiang; Liu, Baohui; Kanazawa, Akira;Yamada, Tetsuya; Abe, Jun

Citation Frontiers in plant science, 9, 1867https://doi.org/10.3389/fpls.2018.01867

Issue Date 2019-01-08

Doc URL http://hdl.handle.net/2115/73066

Rights

© 2019 Zhu, Takeshima, Harigai, Xu, Kong, Liu, Kanazawa, Yamada and Abe. This is an open-access articledistributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution orreproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited andthat the original publication in this journal is cited, in accordance with accepted academic practice. No use, distributionor reproduction is permitted which does not comply with these terms.

Rights(URL) http://creativecommons.org/licenses/by/4.0/

Type article

File Information fpls-09-01867.pdf

Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP

fpls-09-01867 December 28, 2018 Time: 17:59 # 1

ORIGINAL RESEARCHpublished: 08 January 2019

doi: 10.3389/fpls.2018.01867

Edited by:Anna Maria Mastrangelo,

Centro di Ricerca per l’Orticoltura(CRA), Italy

Reviewed by:Humira Sonah,

Laval University, CanadaRichard Macknight,

University of Otago, New ZealandFang Huang,

Nanjing Agricultural University, China

*Correspondence:Baohui Liu

liubh@neigaehrb.ac.cnJun Abe

jabe@res.agr.hokudai.ac.jp

Specialty section:This article was submitted to

Plant Breeding,a section of the journal

Frontiers in Plant Science

Received: 27 September 2018Accepted: 04 December 2018

Published: 08 January 2019

Citation:Zhu J, Takeshima R, Harigai K,

Xu M, Kong F, Liu B, Kanazawa A,Yamada T and Abe J (2019) Lossof Function of the E1-Like-b Gene

Associates With Early FloweringUnder Long-Day Conditions

in Soybean. Front. Plant Sci. 9:1867.doi: 10.3389/fpls.2018.01867

Loss of Function of the E1-Like-bGene Associates With EarlyFlowering Under Long-DayConditions in SoybeanJianghui Zhu1, Ryoma Takeshima2, Kohei Harigai1, Meilan Xu1,3, Fanjiang Kong3,4,Baohui Liu4* , Akira Kanazawa1, Tetsuya Yamada1 and Jun Abe1*

1 Research Faculty of Agriculture, Hokkaido University, Sapporo, Japan, 2 Institute of Crop Science, National Agricultureand Food Research Organization, Tsukuba, Japan, 3 Key Laboratory of Soybean Molecular Design Breeding, NortheastInstitute of Geography and Agroecology, Chinese Academy of Sciences, Harbin, China, 4 School of Life Sciences,Guangzhou University, Guangzhou, China

Photoperiod response of flowering determines plant adaptation to different latitudes.Soybean, a short-day plant, has gained the ability to flower under long-dayconditions during the growing season at higher latitudes, mainly through dysfunctionof phytochrome A genes (E3 and E4) and the floral repressor E1. In this study, weidentified a novel molecular genetic basis of photoperiod insensitivity in Far-EasternRussian soybean cultivars. By testcrossing these cultivars with a Canadian cultivarHarosoy near-isogenic line for a recessive e3 allele, followed by association tests and finemapping, we determined that the insensitivity was inherited as a single recessive genelocated in an 842-kb interval in the pericentromeric region of chromosome 4, whereE1-Like b (E1Lb), a homoeolog of E1, is located. Sequencing analysis detected a single-nucleotide deletion in the coding sequence of the gene in insensitive cultivars, whichgenerated a premature stop codon. Near-isogenic lines (NILs) for the loss-of-functionallele (designated e1lb) exhibited upregulated expression of soybean FLOWERINGLOCUS T (FT ) orthologs, FT2a and FT5a, and flowered earlier than those for E1Lbunder long-day conditions in both the e3/E4 and E3/E4 genetic backgrounds. TheseNILs further lacked the inhibitory effect on flowering by far-red light–enriched long-day conditions, which is mediated by E4, but not that of red-light–enriched long-dayconditions, which is mediated by E3. These findings suggest that E1Lb retards floweringunder long-day conditions by repressing the expression of FT2a and FT5a independentlyof E1. This loss-of-function allele can be used as a new resource in breeding ofphotoperiod-insensitive cultivars, and may improve our understanding of the functionof the E1 family genes in the photoperiod responses of flowering in soybean.

Keywords: soybean, Glycine max, flowering, E1Lb, photoperiodism, adaptation

INTRODUCTION

Photoperiod response of flowering determines the adaptation of crops to a wide range of latitudeswith different daylengths during growing seasons. Its regulatory mechanisms vary with plantspecies, and may rely on both evolutionally conserved and species-specific gene systems. InArabidopsis, a long-day (LD) plant,CONSTANS (CO) plays a key role in regulation of photoperiodic

Frontiers in Plant Science | www.frontiersin.org 1 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 2

Zhu et al. E1Lb Represses Flowering Under LD

flowering; transcriptional and post-translational regulation of COresults in accumulation of the CO protein in the late afternoonunder LD conditions, which in turn activates FLOWERINGLOCUS T (FT) florigen gene expression (reviewed by Andrés andCoupland, 2012; Song et al., 2013). Similarly, in rice, a short-day(SD) plant, a CO ortholog, Heading date 1 (Hd1) (Yano et al.,2000), regulates the FT orthologs Heading date 3a (Hd3a) andRice FT-like 1 (RFT1) (Kojima et al., 2002; Tamaki et al., 2007).However, unlike in Arabidopsis, Hd1 activates Hd3a expressionunder inductive SD conditions, but suppresses it under non-inductive LD conditions (Izawa et al., 2002). This functionalswitch, which is absent in Arabidopsis, is controlled by a complexof Hd1 with the monocot-specific CCT domain protein Grainnumber, plant height and heading date 7 (Ghd7) (Xue et al.,2008); Ghd7 represses the expression of the B-type responseregulator Early heading date 1 (Ehd1) (Doi et al., 2004), anactivator of Hd3a and RFT1 expression, by binding to its cis-regulatory region (Nemoto et al., 2016).

Soybean (Glycine max) has multiple CO orthologs (Fan et al.,2014; Wu et al., 2014), of which two pairs of homoeologs,CO-like (COL) 1a/COL1b and COL2a/COL2b, fully complementthe function of CO in Arabidopsis (Wu et al., 2014). COL1aoverexpression in soybean causes late flowering, and artificialCOL1b mutants flower significantly earlier than the wild type,indicating that both COL1a and COL1b function as floralsuppressors under LD conditions, as in rice (Cao et al., 2015).However, unlike in the case of Hd1, the overexpression of COL1adoes not promote flowering under inductive SD conditions,although it up-regulates major soybean FT orthologs, FT2a andFT5a (Kong et al., 2010; Cao et al., 2015).

Despite the conserved roles of CO and COL genes across plantspecies in photoperiodic flowering, there is no report that anyCOL genes are involved in the genetic variation of floweringtime in soybean. Among the 11 major genes for floweringthat have been reported so far (E1–E9 and J, reviewed by Caoet al., 2017; E10, Samanfar et al., 2017), four maturity genes,E1 to E4, are the main contributors to soybean adaptation to awide range of latitudes (Liu et al., 2011; Jia et al., 2014; Jianget al., 2014; Langewisch et al., 2014; Tsubokura et al., 2014;Zhai et al., 2014; Lu et al., 2015; Kurasch et al., 2017; Li et al.,2017). The floral repressor E1 encodes a protein that containsa bipartite nuclear localization signal and a region distantlyrelated to the B3 domain, and is a possible transcription factorthat represses FT2a and FT5a expression (Xia et al., 2012).E1 expression is up-regulated under LD conditions under thecontrol of the phytochrome A (phyA) proteins E3 and E4 (Liuet al., 2008; Watanabe et al., 2009; Xia et al., 2012). E2, asoybean ortholog of Arabidopsis GIGANTEA (GI) (Watanabeet al., 2011), inhibits flowering under LD conditions througha pathway distinct from the phyA-regulated E1 pathway (Xuet al., 2015; reviewed by Cao et al., 2017). E1 has two homologs,E1-like-a (E1La) and E1Lb, encoded 10,640 kb apart fromeach other in the homoeologous region of chromosome 4 (Xiaet al., 2012; Xu et al., 2015). Down-regulation of the E1Lgenes by virus-induced gene silencing (VIGS) in a cultivardeficient in the E1 gene leads to early flowering and abolishesthe night-break response, suggesting that the two E1L genes

are also involved in the photoperiod responses of soybean(Xu et al., 2015).

Photoperiod insensitivity in soybean is conditioned bycombinations of various alleles at E1, E3, and E4 (Tsubokuraet al., 2013, 2014; Xu et al., 2013; Zhai et al., 2015). E3 and E4were originally identified as major genes for different responsesof flowering to artificially induced LD conditions, where naturaldaylength was extended to 20 h with red light (R)-enriched coolwhite fluorescent lamps (fluorescent-long daylength; FLD) orfar red light (FR)-enriched incandescent lamps (incandescent-long daylength; ILD) (Buzzell, 1971; Buzzell and Voldeng, 1980;Saindon et al., 1989). e3 conditions flowering under the FLDcondition (Buzzell, 1971), whereas e4 does so under the ILDcondition in the e3 background (Saindon et al., 1989), suggestingthat E3 and E4 are functionally diverged and have an epistaticrelationship. On the basis of the functions of alleles at the threeloci, Xu et al. (2013) classified ILD-insensitive cultivars into threegenotypic groups: (group 1) the dysfunction of both E3 and E4;(group 2) the dysfunction of E1 in combination with that of eitherE3 or E4; and (group 3) a combination of e1-as (hypomorphicallele), e3, and E4. Because E4 inhibits flowering under ILDconditions (Saindon et al., 1989; Cober et al., 1996; Abe et al.,2003; Liu and Abe, 2010), the group 3 cultivars have novel genesthat abolish or reduce ILD sensitivity. One such gene is an early-flowering allele at qDTF-J, a QTL for days to flowering in linkagegroup J, which encodes FT5a; early flowering is caused by itsincreased transcriptional activity or mRNA stability associatedwith an insertion in the promoter and/or deletions in the 3′UTR (Takeshima et al., 2016). Here, we describe a novel loss-of-function allele at the E1Lb locus, which is most likely involvedin the gain of photoperiod insensitivity in group 3 soybeancultivars. Our data suggest that E1Lb inhibits flowering underLD conditions, independently of E1, and play major roles in thecontrol of flowering in soybean.

MATERIALS AND METHODS

Plant Materials and Segregation AnalysisThe indeterminate Far-Eastern Russian soybean cultivars Zeika(ZE), Yubileinaya (YU), and Sonata were crossed with theCanadian indeterminate cultivar Harosoy (L58-266; HA); ZE andYU were also crossed with a Harosoy near-isogenic line (NIL)for e3 (PI547716; H-e3). The three Russian cultivars have thesame genotype as H-e3 at five maturity loci, E1, E2, E3, E4,and E9 (e1-as/e2/e3/E4/E9), but unlike H-e3 they flower withoutany marked delay under ILD conditions in comparison withnatural daylength (ND) conditions (maximum daylength, 15.2 h)in Sapporo, Japan (43◦07′N, 141◦35′E) (Xu et al., 2013). TheILD condition was set at an experimental farm of HokkaidoUniversity by extending the ND to 20 h by supplemental lightingfrom 2:00 to 7:00 and from 18:00 to 22:00 with incandescentlamps with a red-to-far-red (R:FR) quantum ratio of 0.72 (Abeet al., 2003). Seeds of F2 populations and parents were sown inpaper pots (Paperpots No. 2, Nippon Beet Sugar ManufacturingCo., Tokyo, Japan) on 28 May 2013 for the crosses with HA and26 May 2014 for crosses with H-e3. The pots were put under the

Frontiers in Plant Science | www.frontiersin.org 2 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 3

Zhu et al. E1Lb Represses Flowering Under LD

ILD condition, and 12 days later seedlings were transplanted intosoil. The progeny test was carried out for 48 F2 plants randomlyselected from the H-e3 × ZE cross and recombinant plants usedin fine mapping. Seeds of these plants were sown in paper potson late May in 2015 to 2017 (25 May, 2015; 28 May, 2016; and 26May, 2017). After 12 days under the ILD condition, 15 seedlingsper plant were transplanted into the same field. The number ofdays from sowing to the first flower opening (R1) (Fehr et al.,1971) of each plant was recorded.

Association Test, Linkage MapConstruction, and Fine MappingA total of 16 F2 plants from the H-e3 × ZE cross were usedto test the association of ILD sensitivity with simple sequencerepeat (SSR) marker genotypes. They were selected based on thesegregation pattern in their progeny, and included 8 plants fixedfor ILD insensitivity and 8 plants fixed for ILD sensitivity. SSRmarkers were chosen from those located in genomic regions thatharbored the soybean orthologs of Arabidopsis flowering genes(Song et al., 2004; Watanabe et al., 2012). The SSR markerssignificantly associated with ILD sensitivity were genotyped fora total of 306 F2 plants from the H-e3 × ZE and H-e3 × YUcrosses to confirm the detected association. Plants recombinantin the targeted region were subjected to fine mapping; thegenotypes for the target gene were estimated based on thesegregation of flowering under the ILD condition in the progenyand were compared with the graphical genotypes constructed byusing additional 11 BARCSOY SSR markers (Song et al., 2010)(Supplementary Table S1).

Development of NILsFour sets of NILs, each including one NIL for ILD insensitivityand another one for sensitivity, were developed fromheterozygous inbred F5 plants derived from different F2plants (#4 and #21) from the H-e3 × ZE cross and those (#11and #20) from the HA × ZE cross. The former two sets ofNILs had the recessive e3 allele, whereas the latter two had thedominant E3 allele. These lines, together with parents and anILD-insensitive NIL of HA for e3 and e4 (PI546043; H-e3e4),were cultivated in a growth chamber (25◦C, 20-h daylength)with an average photon flux of 120 µmol m−2 s−1 and an R:FRratio of 2.2 at 1 m below light sources, or in the field under theILD condition (sowing date, May 26, 2018), as described above.For comparison, indeterminate NILs for alleles, e1-nl and e1-as,at E1 (NIL-E1; e2/E3/E4/E9), which were developed from aheterozygous inbred F5 plant derived from a cross between theJapanese determinate cultivar Toyomusume (e1-nl/e2/E3/E4/e9)and HA, were included in the evaluation of flowering under theILD condition.

DNA Extraction and SSR Marker AnalysisTotal DNA was extracted from trifoliate leaves of each of 150 H-e3 × ZE and 156 H-e3 × YU F2 plants as described by Doyleand Doyle (1990), and from each of 492 seeds from two F2 plantsfrom the H-e3× ZE cross, as described by Xia et al. (2012). EachPCR mixture for SSR marker analysis contained 30 ng of total

genomic DNA as a template, 0.2 µl of each primer (10 µM), 0.8 µlof dNTPs (2.5 mM), 0.1 µl of Taq DNA polymerase (Ampliqon),and 1 µl of 10× ammonium buffer (Ampliqon) in a total volumeof 10 µl; amplification conditions were 35 cycles at 94◦C for 30 s,55◦C for 30 s, and 72◦C for 30 s. PCR products were separated byelectrophoresis in 10.5% (w/v) polyacrylamide gels, stained withethidium bromide, and visualized under UV light.

Expression AnalysisA new fully expanded leaflet was sampled from each of fourplants per parent and NIL at Zeitgeber time 3 in two differentgrowing stages, the 2nd and 3rd leaf stages. The sampled leaveswere bulked, immediately frozen in liquid N2, and stored at−80◦C. Total RNA was isolated from frozen tissues with TRIzolReagent (Thermo Fisher Scientific). DNase I (Takara) was used toremove genomic DNA. The complementary DNAs (cDNAs) weresynthesized from 1 µg of total RNA by using the M-MLV reversetranscriptase system (Invitrogen) with an oligo (dT) 20 primerin a volume of 20 µL. Transcript levels of E1, E1La, E1Lb, FT2a,and FT5a were determined by quantitative real-time PCR. ThePCR mixture (20 µL) contained 0.1 µL of the cDNA synthesisreaction mixture, 5 µL of 1.2 µM primer premix, and 10 µLSYBR Premix Ex Taq II (Takara). A CFX96 Real-Time System(Bio-Rad) was used. The PCR cycling conditions were 95◦C for3 min followed by 40 cycles of 95◦C for 10 s, 59◦C for 30 s,72◦C for 20 s, and 78◦C for 2 s. Fluorescence was quantifiedbefore and after the incubation at 78◦C to monitor the formationof primer dimers. The mRNA for β-tubulin was used fornormalization. A reaction mixture without reverse transcriptasewas also used as a control to confirm the absence of genomicDNA contamination. Amplification of a single DNA fragmentwas confirmed by melting curve analysis and gel electrophoresisof the PCR products. Averages and standard errors of relativeexpression levels were calculated from PCR results for threeindependently synthesized cDNAs. Primer sequences used inexpression analyses are listed in Supplementary Table S1.

Sequencing and Marker Analysis of E1LbThe coding sequences of the three gene models,Glyma.04G143300, Glyma.04G143400 and Glyma.04G143500,were analyzed for H-e3 and ZE. The coding sequenceswere amplified from the cDNAs by using primers listed inSupplementary Table S1. The amplified fragments were clonedinto a pGEM-T Easy vector (Promega) and sequenced witha BigDye Terminator v3.1 Cycle Sequencing kit and an ABIPRISM 3100 Avant Genetic Analyzer (both from AppliedBiosystems, Japan) according to the manufacturer’s instructions.A derived cleaved amplified polymorphic sequence (dCAPS)marker targeting a single-base deletion observed in ZE wasdeveloped to discriminate the functional E1Lb allele of H-e3from the loss-of-function e1lb allele of ZE. The 275-bp DNAfragment amplified from ZE by PCR with the forward primer5′-GTGTAAACACTCAAAGTCCTT-3′ and the reverse primer5′-CGTCTTCTTGATCTTCCAACG-3′ was digested withHpyCH4IV (New England Biolabs Japan) into two fragments,254 bp and 21 bp, but the 276-bp fragment amplified fromH-e3 was resistant to HpyCH4IV digestion. The PCR products

Frontiers in Plant Science | www.frontiersin.org 3 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 4

Zhu et al. E1Lb Represses Flowering Under LD

were treated with HpyCH4IV for 1 h and then separated byelectrophoresis in 2.5% NuSieve 3:1 gel (Lonza), stained withethidium bromide, and visualized under UV light.

Survey of the Dysfunctional Allele inILD-Insensitive AccessionsA total of 62 ILD-insensitive accessions including the threeRussian cultivars were surveyed for the E1Lb genotype usingthe allele-specific DNA marker. They included 9 accessionsfrom northern Japan, 26 from north-eastern China, 16 fromFar-Eastern Russia, 8 from Ukraine, and 3 from Poland(Supplementary Table S2). The maturity genotypes at E1 to E4of 50 accessions were determined previously by Xu et al. (2013),and those of the remaining 12 accessions were assayed accordingto Xu et al. (2013) and Tsubokura et al. (2014).

RESULTS

Segregation of Flowering Time in F2 andF3 PopulationsThe three Russian cultivars are photoperiod insensitive (Xu et al.,2013). They flowered 45–47 days after sowing (DAS) underthe ND condition of Sapporo, whereas H-e3 and HA floweredapproximately 5 and 10 days later, respectively. Under the ILDcondition, the three cultivars and H-e3 flowered 2–4 days andaround 20 days later than under ND, respectively, whereas HAcontinued vegetative growth and did not develop any flower budsuntil the end of light supplementation (10 August, 76 DAS).

Flowering time under the ILD condition in F2 populationsof the H-e3 × ZE and H-e3 × YU crosses varied continuouslyfrom that of ILD-insensitive parents (45 DAS for ZE and 46 DASfor YU) to the end of light supplementation; 10 out of 150 and12 out of 156 plants had no flower buds in the H-e3 × ZEand H-e3 × YU F2 populations, respectively (Figure 1). In

both populations, the distribution of flowering time tended tobe bi-modal; plants which flowered at 56 DAS and later orremained vegetative segregated more than those which floweredearlier. We randomly selected 48 H-e3 × ZE F2 plants andtested their progeny for flowering time segregation under theILD condition. Based on the segregation pattern, the 48 F2plants could be classified into three groups: (1) plants fixedfor ILD insensitivity (all F3 plants tested flowered as ZE did;e/e); (2) those segregating for flowering time (E/e) and (3)those fixed for ILD sensitivity (all F3 plants tested showeddelayed or no flowering; E/E) (Figure 1A). The number ofplants was 8 in e/e, 23 in E/e, and 17 in E/E, in consistencewith a monogenic 1:2:1 ratio (χ2 = 3.81, df = 2, p = 0.18),suggesting the involvement of a single recessive gene for ILDinsensitivity. Based on the results of the progeny test, we classified306 F2 plants into early-flowering ILD-insensitive plants, whichflowered before 56 DAS, and late- or non-flowering ILD-sensitiveplants (Figure 1). The segregation ratios of the two classes fit theexpected 3:1 ratio (χ2 = 0.33, df = 1, p = 0.56 for H-e3× ZE,χ2 = 3.28, df = 1, p = 0.07 for H-e3 × YU), confirmingthat ILD insensitivity is controlled mainly by a single recessivegene.

We also examined the segregation of flowering time underthe ILD condition for the crosses between HA and the threeRussian cultivars. Because HA had the E3 allele and the threecultivars had the e3 allele, we predicted that, in addition tothe gene for ILD insensitivity segregated in the crosses with H-e3, the E3 locus would also segregate in the F2 populations.In the three crosses, however, ILD-insensitive plants segregatedat frequencies of 21.1–33.9%; the remaining plants remainedvegetative until the end of light supplementation (Table 1).These segregation frequencies were thus inconsistent withthose of a two-gene model, but were close to those expectedfrom monogenic inheritance, as in the crosses with H-e3(Table 1).

FIGURE 1 | Segregation of flowering time in F2 populations of crosses between a Harosoy NIL for e3 (H-e3) and the incandescent-long daylength (ILD)-insensitivecultivars Zeika (ZE) and Yubileinaya (YU) under far red light–enriched ILD conditions. (A) H-e3 × ZE; (B) H-e3 × YU. In a cross between H-e3 and ZE, 48 F2 plantswere selected for the progeny test; ILD-sensitivity genotypes were estimated based on the segregation in the progeny. Pink bars, homozygotes for ILD insensitivity(e/e); yellow–green bars, heterozygotes (E/e); light-blue bars, homozygotes for ILD sensitivity (E/E). Arrows indicate mean values of flowering time in parents. Dottedvertical lines indicate the threshold for classification of F2 plants into early-flowering ILD-insensitive and late- or non-flowering ILD-sensitive. nf, no flower buds by theend of light supplementation. DAS, days after sowing.

Frontiers in Plant Science | www.frontiersin.org 4 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 5

Zhu et al. E1Lb Represses Flowering Under LD

TABLE 1 | Segregation of ILD-insensitivity in F2 of crosses of an ILD-sensitive cultivar Harosoy with ILD-insensitive Russian cultivars.

Cross combination Number of plants χ2 value for 1:3 P-value

ILD-insensitive ILD-sensitive Total

Harosoy × Zeika 19 37 56 3.57 0.059

Harosoy × Yubileinaya 28 105 133 1.66 0.198

Harosoy × Sonata 19 54 73 0.06 0.803

FIGURE 2 | Simple sequence repeat (SSR) marker analyses in F2 plants from a Harosoy isoline for e3 (H-e3) × Zeika (ZE) cross. (A) Gel electrophoresis for theanalysis of Satt190 and Sat_085. Eight plants homozygous for the ILD-insensitive allele (e/e) and 8 plants homozygous for the ILD-sensitive allele (E/E) were selectedon the basis of the results of the progeny test. M1, ϕX174/HaeIII digest; M2, 100 bp DNA ladder. (B,C) Association between Sat_085 and flowering time inH-e3 × ZE (B) and H-e3 × YU (C). F2 plants were classified based on the genotype at Sat_085. Pink bars, homozygotes for the allele from ILD-insensitive parents(I/I); yellow-green bars, heterozygotes (I/S); light-blue bars, homozygotes for the allele from ILD-sensitive H-e3 (S/S). Arrows indicate mean values of flowering time inparents. DAS, days after sowing.

Association Test, Linkage MapConstruction, and Fine MappingTo determine the genomic position of the gene for ILDinsensitivity from ZE, we tested the association between ILDsensitivity and SSR marker genotypes. Based on the results of theprogeny test, we selected 16 F2 plants from the H-e3× ZE cross,8 homozygous for ILD insensitivity (e/e), and 8 homozygous forILD sensitivity (E/E). Among the SSR markers tested, Satt190and Sat_085 in linkage group C1 (chromosome 4; Chr04)showed genotypic variation in complete accordance with the ILDsensitivity (Figure 2A). Then we determined the genotypes ofthe two markers in the whole F2 plants of H-e3 × ZE and H-e3 × YU populations (Figures 2B,C). The two markers weretightly linked to each other with a recombination value of 2.1,and were closely associated with ILD sensitivity. All of theplants homozygous for the allele from ILD-insensitive parents atSat_085 (I/I) flowered before 56 DAS (H-e3 × ZE) or 52 DAS(H-e3 × YU), whereas those homozygous for the allele fromILD-sensitive H-e3 (S/S) flowered at ≥60 DAS or did not flower

in both crosses. Heterozygous plants (I/S) mostly flowered at≥58 DAS (H-e3 × ZE) or ≥54 DAS (H-e3 × YU), which partlyoverlapped with the flowering date ranges of the S/S plants; onlya few plants flowered as early as the I/I plants. These resultsstrongly suggested that a gene for ILD insensitivity is located nearthe two SSR markers.

Satt190 and Sat_085 are located 17.3 Mb from each otherin the pericentromeric region of Chr04 (Schmutz et al., 2010)(Phytozome v12.1/Glycine max Wm82.a2.v1). To delimitthe genomic region of the gene for ILD insensitivity moreprecisely, we selected plants with recombination betweenthe two markers (7 from 306 F2 plants from the H-e3× ZEand H-e3 × YU crosses and 3 from 492 F3 plants from theH-e3 × ZE cross) and constructed their graphical genotypeswith 11 SSR markers. A comparison of the graphical genotypeswith the genotype of ILD insensitivity estimated by the progenytest revealed that the gene for ILD insensitivity was locatedbetween SSR markers M5 (BARC-18g-0889) and M6 (BARC-18g-0895) (Figure 3). The physical distance between the two

Frontiers in Plant Science | www.frontiersin.org 5 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 6

Zhu et al. E1Lb Represses Flowering Under LD

FIGURE 3 | Fine mapping of a locus for ILD insensitivity (E/e) and annotated genes in the delimited region of Williams 82 chromosome 4 (Glycine max Wm82.a2.v1).M1 to M10, BARCSOY-SSR markers; Pink bars, homozygotes for the allele from ILD-insensitive parents (I/I); yellow–green bars, heterozygotes (I/S); light blue bars,homozygotes for the allele from ILD-sensitive H-e3 (S/S). The E genotype for ILD insensitivity was estimated from the segregation patterns in the progeny. Six genesannotated in the region between M5 and M6 are shown at the bottom.

markers was 842 kb, and the delimited region contained only 6annotated genes (Phytozome v12.1/Glycine max Wm82.a2.v1)(Figure 3 and Table 2). RNA-sequencing Atlas in Phytozomev12.1/Glycine maxWm82.a2.v1 indicates that Glyma.04G143000,Glyma.04G143100 and Glyma.04G143200 are expressedonly in flower or root tissues, whereas Glyma.04G143300,Glyma.04G143400, and Glyma.04G143500 are expressed inleaves (Severin et al., 2010). Because ZE exhibited significantlyhigher expressions for FT2a and FT5a in leaves in the 2ndand 3rd trifoliate leaf stages than H-e3 under R-enriched LDcondition (Supplementary Figure S1), we focused on thethree genes expressed in leaves as a possible candidate ofthe gene for ILD insensitivity that upregulates the two FTgenes.

Sequence AnalysisSequence analysis revealed that ZE and H-e3 possessed identicalsequences for Glyma.04G143400 and Glyma.04G143500,whereas one of cytosines at the 162th nucleotide to 164thnucleotide from the adenine of the start codon was deleted in theGlyma.04G143300 from ZE; this deletion generated a prematurestop codon, and the Glyma.04G143300 from ZE was predictedto encode a truncated protein of 61 amino acids (Figure 4).Glyma.04G143300 is E1Lb, one of two homoeologs (E1La and

E1Lb) of floral repressor E1 (Xia et al., 2012). Because the down-regulation of E1La and E1Lb expressions by VIGS promotesflowering under non-inductive conditions such as LD and nightbreak (Xu et al., 2015), we considered the loss-of-function alleleof E1Lb (designated e1lb hereafter) as the most probable causalfactor for the ILD-insensitivity.

TABLE 2 | Genes annotated in an 842-kb genomic region in chromosome 4delimited by fine-mapping.

No. Gene Annotation (PhytozomeV12.1/Glycine max Wm82.a2.v1)

Expressedtissues

(1) Glyma.04G143000 Diacylglycerol kinase 7 Flower

(2) Glyma.04G143100 RNA-binding (RRM/RBD/RNPmotifs) family protein

Root

(3) Glyma.04G143200 Pectin lyase-like superfamily protein Flower

(4) Glyma.04G143300 AP2/B3-like transcriptional factorfamily protein, E1Lb

Leaf

(5) Glyma.04G143400 Cytidine/deoxycytidylate deaminasefamily protein

Leaf, root

(6) Glyma.04G143500 Mitochondrial substrate carrierfamily protein

Flower, leaf

Data on expressed tissues are referred from Glycine max Wm82.a2.v1. (Severinet al., 2010).

Frontiers in Plant Science | www.frontiersin.org 6 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 7

Zhu et al. E1Lb Represses Flowering Under LD

FIGURE 4 | DNA and predicted amino acid sequences of Glyma.04G143300 (E1Lb) in Williams 82 (W82), Harosoy isoline for e3 (H-e3), and Zeika (ZE). (A) DNAsequences of the 138th nucleotide to 177th nucleotide from the adenine of the stat codon. One of cytosines at the 162th nucleotide to 164th nucleotide underlinedwas deleted in ZE. (B) Predicted amino acid sequences.

We developed a dCAPS marker to discriminate e1lb fromE1Lb (Figure 5). The PCR-amplified fragment of 275 bp fromZE produced a shorter fragment of 254 bp when digestedwith HpyCH4IV, whereas that from H-e3 (276 bp) was notdigested. The digestion of the PCR products from YU and Sonata(Russian cultivar) produced 254-bp fragments, indicating thatthese two cultivars had the same deletion as ZE (Figure 5B).Therefore, the segregation of ILD-insensitive plants in the crossesof these cultivars with HA and H-e3 were most likely caused bye1lb.

Comparison of Flowering Time and GeneExpression Among NILsWe evaluated the allelic effects of E1Lb and e1lb on floweringunder the R-enriched LD condition (daylength, 20 h) in four setsof NILs, each for E1Lb and e1lb, developed from different F2plants from the H-e3 × ZE cross (#4 and #21) and the HA × ZEcross (#11 and #20). In the two sets of the e3/E4 NILs, eachNIL for e1lb flowered at the same or almost the same time(#4, 31.7 DAS; #21, 30.3 DAS) as ZE (30.3 DAS); this was on

average 6.7–7.6 days earlier than the respective NILs for E1Lb,which flowered at almost the same time as H-e3 (Figure 6A).Flowering times of the E3/E4 NILs were around 20 days ormore later than those of the e3/E4 NILs. e1lb also promotedflowering in the E3/E4 background: each NIL for e1lb floweredaround 10 days earlier than the respective NIL for E1Lb andHA. This flowering-promoting effect of e1lb versus E1Lb underthe R-enriched LD condition was smaller than that of e4 vs.E4 and that of e3 vs. E3, because H-e3e4 and H-e3 flowered,on average, 13 and 25 days earlier than H-e3 and HA (E3E4),respectively.

We also evaluated the effect of e1lb vs. E1Lb on floweringunder the FR-enriched ILD condition (Figure 6B). e1lb inducedflowering at 58 DAS (#4) or 49 DAS (NILs #21) in the e3/E4genetic background and at 56 DAS (#11 and #20) in the E3/E4genetic background. All these NILs produced pods of up to3 cm in length at the end of light supplementation, similarto those of ZE and H-e3e4. In contrast, the e3/E4 NILs forE1Lb and H-e3 flowered around 20 days later, and E3/E4 NILsfor E1Lb and HA continued vegetative growth and did notproduce any flower buds until the end of light supplementation.

Frontiers in Plant Science | www.frontiersin.org 7 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 8

Zhu et al. E1Lb Represses Flowering Under LD

FIGURE 5 | Derived cleaved amplified polymorphic sequence (dCAPS) marker analysis to detect the single-base deletion in the e1lb allele. (A) Primers designed andthe generation of an HpyCH4IV restriction site. One of three cytosines marked by red was deleted in ZE. (B) Gel electrophoresis of PCR products without (–) or with(+) HpyCH4IV digestion. H-e3, Harosoy NIL for e3; ZE, Zeika.

Therefore, e1lb was sufficient to induce flowering under theILD condition, irrespective of the E3 genotype (Figure 6B).Interestingly, a similar flowering-promoting effect was observedin the NIL-E1 for a loss-of-function allele e1-nl (e1); itinitiated flowering under the ILD condition, as the E3/E4NILs for e1lb, whereas the NIL for e1-as (E1) did not(Figure 6B).

We tested the expression levels of E1, two E1L genes, andtwo FT orthologs in two different growing stages (the 2nd and3rd leaf stages) in the e3/E4 NILs grown under the R-enrichedLD condition (Figure 7). The expression levels of E1 and E1Lawere similar between the NILs for E1lb and e1lb at both stagesin NILs #4 or at the 3rd stage in NILs #21; both E1 and E1Lawere significantly up-regulated in the 2nd leaf stage in NILs(#21) for e1lb relative to those for E1Lb. On the other hand,the expression of E1Lb was significantly down-regulated in theNILs for e1lb at both stages (#4) or at the 3rd leaf stage (#21).In contrast, the expression of both FT2a and FT5a was up-regulated at both stages in the NILs for e1lb relative to thosefor E1Lb in both NIL sets. The similar effect of e1lb vs. E1Lb onthe expression of FT2a and FT5a was also observed at the 3rdleaf stage in both sets of E3/E4 NILs (#11 and #20; Figure 8).As observed in the e3/E4 NILs for e1lb, the expression levels ofFT2a and FT5a were significantly upregulated in the E3/E4 NILfor e1lb.

Survey of the e1lb Allele inILD-Insensitive Soybean AccessionsTo determine whether or not the deletion in the E1Lb gene isregion specific, we surveyed polymorphism in the ILD-insensitivesoybean accessions analyzed for the genotypes of E1 to E4 byusing the developed dCAPS marker. In addition to the threeRussian cultivars, we found that another two Russian cultivars,Salyut 216 and DYA-1, had the e1lb allele, whereas all theother accessions had the functional E1Lb allele (SupplementaryTable S2). All of Russian cultivars with e1lb possessed thematurity genotype of e1-as/e3/E4. There was no cultivar whichhas loss-of-function alleles at both E1 and E1Lb loci.

DISCUSSION

The soybean maturity loci, E1 to E4, are major flowering locithat determine the ability of cultivars to adapt to differentlatitudes. Diverse allelic combinations at the E1, E3, and E4 loci,each of which has multiple loss-of-function alleles (Tsubokuraet al., 2013, 2014; Xu et al., 2013; Jiang et al., 2014; reviewedby Cao et al., 2017), have resulted in cultivars with varioussensitivities to photoperiod. Photoperiod insensitivity is anadaptive trait for cultivars at high latitudes; such cultivars areclassified into three genotypic groups according to the allelic

Frontiers in Plant Science | www.frontiersin.org 8 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 9

Zhu et al. E1Lb Represses Flowering Under LD

FIGURE 6 | Flowering time in NILs for the e1lb (e) and E1Lb (E) alleles under R-enriched and FR-enriched LD conditions. Two sets of NILs (#4 and #21) have thee3/E4 genotype, similar to H-e3, whereas the other two (#11 and #20) have the E3/E4 genotype, similar to HA. A set of E3/E4 NILs for e1-nl (e1) and e1-as (E1) atE1 locus (NIL-E1) were also evaluated for the flowering under the FR-enriched LD condition. Plants were grown under 20-h (A) R-enriched LD or (B) FR-enriched ILDconditions. Data are presented as mean and standard error (n = 5). nf, no flower buds by the end of light supplementation, DAS, days after sowing, ∗∗∗p < 0.001.

functions at each of the three loci (Xu et al., 2013). Among theILD accessions tested, the predominant group has the loss-of-function alleles of the phyA genes E3 and E4 (e3/e4), followedby a group which has the loss-of-function of the E1 repressorfor FT2a and FT5a in combination with a dominant E3 orE4 allele. Cultivars of the third group have a novel geneticmechanism that abolishes or reduces sensitivity to daylength,because they have the same genotype (e1-as/e3/E4) as an HANIL for e3, which is sensitive to FR-enriched ILD conditions(Saindon et al., 1989; Cober et al., 1996; Abe et al., 2003;Liu and Abe, 2010; Xu et al., 2013). Takeshima et al. (2016)carried out QTL analysis of ILD insensitivity by a testcross ofa Chinese cultivar of group 3 with the HA NIL for e3 anddemonstrated that an early-flowering allele at qDTF-J, whichencodes the FT5a protein, up-regulates FT5a expression by cis-activation in the presence of E4 to induce flowering under ILDconditions.

In the present study, we detected a novel loss-of-functionallele that resulted from a frameshift mutation at the E1Lblocus in Far-Eastern Russian group 3 photoperiod-insensitivecultivars. E1Lb and its tandemly linked homolog, E1La, have highsequence similarity to E1, suggesting their functional similarity,although a certain degree of subfunctionalization is suggestedby the presence of a number of amino acid substitutions andindels between the E1 and E1L genes (Xia et al., 2012). Down-regulation by VIGS revealed that, similar to E1, E1L genes inhibitflowering under LD and night-break conditions (Xu et al., 2015),but the function of each homolog has remained undetermined.Comparison of NILs for E1Lb and e1lb in this study suggeststhat e1lb promotes flowering under both R-enriched and FR-enriched LD conditions. In particular, the effect of e1lb vs. E1Lbin the FR-enriched LD condition was similar to that of e4 vs. E4,irrespective of the E3 genotype, suggesting that e1lb completelycancels the inhibitory effect of FR-enriched LD on flowering

Frontiers in Plant Science | www.frontiersin.org 9 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 10

Zhu et al. E1Lb Represses Flowering Under LD

FIGURE 7 | Expression levels of E1, E1-Like, and FT genes at the second and third leaf stages in two sets of e3/E4 NILs for the e1lb (e) and E1Lb (E) alleles underR-enriched LD (20 h) conditions. Relative mRNA levels (mean and standard error, n = 3) are expressed as the ratios to β-tubulin transcript levels. ∗p < 0.05,∗∗∗p < 0.005.

FIGURE 8 | Expression levels of E1, E1-Like, and FT genes at the third leaf stage in two sets of E3/E4 NILs for the e1lb (e) and E1Lb (E) alleles under R-enriched LD(20 h) conditions. Relative mRNA levels (mean and standard error, n = 3) are expressed as the ratios to β-tubulin transcript levels. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.005.

modulated by E4. These flowering-promoting effects are mostlikely due to the up-regulation of FT2a and FT5a; their expressionlevels were not associated with the expression levels of E1 andE1La. One likely explanation for this observation is that the totalexpression level and/or activity of E1, E1La and E1lb may beimportant for the repression of FT2a and FT5a expression. Theinduction of flowering by e1lb in the E3/E4 genetic backgroundunder ILD conditions is in good accordance with monogenic

segregation observed in the crosses of HA with RussianILD-insensitive cultivars. e1lb also promoted flowering underR-enriched LD conditions, but its effect was small and it couldnot cancel flowering inhibition by E3 as efficiently as e3 did. Thefunction of E1Lb may therefore depend on light quality of LD.Interestingly, e1-nl (loss-of-function allele at the E1 locus) couldalso cancel the inhibitory effect of FR-enriched LD conditionson flowering, as efficiently as e4 could. Because the effects of

Frontiers in Plant Science | www.frontiersin.org 10 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 11

Zhu et al. E1Lb Represses Flowering Under LD

e1lb under the e1-as background were similar to those of e1-nl under the E1Lb background (Figure 6B), E1 and E1Lb mayinhibit flowering under LD conditions, independently of eachother. It may be tempting in a further study to develop doublerecessive lines for the loss-of-function alleles at the E1 and E1Lbloci not only to elucidate the interaction between the two genesand the function of another E1 homolog, E1La, but also to explorethe regulatory mechanisms of these E1 family genes by E3 andE4 under different light conditions. In addition, a further studyis also needed to determine why the loss-of-function allele atE1Lb can singly upregulate the FT2a and FT5a expression underLD condition, even though the remaining E1 family genes areexpressed normally.

E1, E2, and E3 have large effects on flowering in a widerange of latitudes, whereas the allelic effect of E4 is ratherlimited to high latitudes (Yamada et al., 2012; Tsubokura et al.,2013, 2014; Xu et al., 2013; Lu et al., 2015). Among thesefour soybean genes, E1 has the most marked effect on time toflowering (McBlain et al., 1987; Upadhyay et al., 1994; Tsubokuraet al., 2014). The polymorphism of E1 (or its flanking genomicregion) largely accounts for the variation in flowering time andrelated agronomic traits in segregating populations of differentgenetic backgrounds (Yamanaka et al., 2000; Wang et al., 2004;Zhang et al., 2004; Funatsuki et al., 2005; Liu et al., 2007; Zhaiet al., 2015). In contrast to the E1 gene, only a few reportshave demonstrated the presence of major genes or QTLs forflowering and maturing associated with the genomic region ofChr04 harboring E1La and E1Lb (Cober et al., 2010; Chenget al., 2011; Watanabe et al., 2017; Kong et al., 2018). Coberet al. (2010) determined that the E8 gene, which was identifiedin a photoperiod-insensitive genetic background, is located in agenomic region harboring two E1L genes 10, 640 kb apart fromeach other (Xu et al., 2015), suggesting either of E1La and E1Lbas a candidate for E8. The QTLs for flowering and maturity werealso detected in the positions of Chr04 similar to that of E8(Cheng et al., 2011; Watanabe et al., 2017; Kong et al., 2018).It would be interesting to determine whether the E8 gene isE1La or E1Lb and to identify the responsible genes for theseQTLs. Genotyping with an allele-specific DNA marker in thisstudy revealed that e1lb is a rare and region-specific allele even inearly maturing photoperiod-insensitive cultivars, suggesting thate1lb has neither largely contributed to the diversity of flowering

behaviors nor been used widely in soybean breeding. The e1lballele may therefore be useful as a new resource to broaden thegenetic variability of soybean cultivars for flowering under LDconditions at high latitudes.

AUTHOR CONTRIBUTIONS

JZ, BL, and JA conducted the experiments. JZ, RT, andKH conducted genetic analyses, fine-mapping, and sequencinganalyses. JZ and MX developed allele-specific DNA markers andanalyzed the variation of genotypes in soybean accessions. JZ,MX, and TY conducted the expression analyses. JZ and JA draftedthe manuscript with edits from RT, KH, MX, FK, BL, TY, and AK.All authors read and approved the final manuscript.

FUNDING

This work was supported in part by Grants-in-Aid for ScientificResearch from the Ministry of Education, Culture, Sports,Science, and Technology of Japan (17K07579) to JA, by theNational Natural Science Foundation of China (Grant No.31501330) to MX, the Program of the General Program ofthe National Natural Science Foundation of China (Grant No.31771815) to BL, and by the Natural Key R&D Program of China(2017YFE0111000 and 2016YFD0100400) to FK.

ACKNOWLEDGMENTS

The authors are grateful to Drs. AY Ala (All Russian ResearchInstitute of Soybean, Russia) and ER Cober (Agriculture andAgri-Food Canada, Canada) for providing us seeds of soybeancultivars and experimental lines.

SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found onlineat: https://www.frontiersin.org/articles/10.3389/fpls.2018.01867/full#supplementary-material

REFERENCESAbe, J., Xu, D., Miyano, A., Komatsu, K., Kanazawa, A., and Shimamoto, Y. (2003).

Photoperiod-insensitive Japanese soybean landraces differ at two maturity loci.Crop Sci. 43, 1300–1304. doi: 10.2135/cropsci2003.1300

Andrés, F., and Coupland, G. (2012). The genetic basis of flowering responses toseasonal cues. Nat. Rev. Genet. 13, 627–639. doi: 10.1038/nrg3291

Buzzell, R. I. (1971). Inheritance of a soybean flowering response to fluorescent-daylength conditions. Can. J. Genet. Cytol. 13, 703–707. doi: 10.1139/g71-100

Buzzell, R. I., and Voldeng, H. D. (1980). Inheritance of insensitivity to longdaylength. Soyb. Genet. Newsl. 7, 26–29. doi: 10.1093/jhered/esp113

Cao, D., Li, Y., Lu, S., Wang, J., Nan, H., Li, X., et al. (2015). GmCOL1aand GmCOL1b function as flowering repressors in soybean under long-dayconditions. Plant Cell Physiol. 56, 2409–2422. doi: 10.1093/pcp/pcv152

Cao, D., Takeshima, R., Zhao, C., Liu, B., Abe, J., and Kong, F. (2017). Molecularmechanisms of flowering under long days and stem growth habit in soybean.J. Exp. Bot. 68, 1873–1884. doi: 10.1093/jxb/erw394

Cheng, L., Wang, Y., Zhang, C., Wu, C., Xu, J., Zhu, H., et al. (2011).Genetic analysis and QTL detection of reproductive period and post-floweringphotoperiod responses in soybean. Theor. Appl. Genet. 123, 421–429. doi: 10.1007/s00122-011-1594-8

Cober, E. R., Molnar, S. J., Charette, M., and Voldeng, H. D. (2010). A new locusfor early maturity in soybean. Crop Sci. 50, 524–527. doi: 10.2135/cropsci2009.04.0174

Cober, E. R., Tanner, J. M., and Voldeng, H. D. (1996). Genetic control ofphotoperiod response in early-maturing, near-isogenic soybean lines. Crop Sci.36, 601–605. doi: 10.2135/cropsci1996.0011183X003600030013x

Doi, K., Izawa, T., Fuse, T., Yamanouchi, U., Kubo, T., Shimatani, Z., et al. (2004).Ehd1, a B-type response regulator in rice, confers short-day promotion of

Frontiers in Plant Science | www.frontiersin.org 11 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 12

Zhu et al. E1Lb Represses Flowering Under LD

flowering and controls FT-like gene expression independently of Hd1. GenesDev. 18, 926–936. doi: 10.1101/gad.1189604

Doyle, J. J., and Doyle, J. L. (1990). Isolation of plant DNA from fresh tissue. Focus12, 13–15.

Fan, C., Hu, R., Zhang, X., Wang, X., Zhang, W., Zhang, Q., et al. (2014). ConservedCO-FT regulons contribute to the photoperiod flowering control in soybean.BMC Plant Biol. 14:9. doi: 10.1186/1471-2229-14-9

Fehr, W. R., Caviness, C. E., Burmood, D. T., and Pennington, J. S. (1971). Stage ofdevelopment descriptions for soybeans, Glycine max (L.) merrill. Crop Sci. 11,929–931. doi: 10.2135/cropsci1971.0011183X001100060051x

Funatsuki, H., Kawaguchi, K., Matsuba, S., Sato, Y., and Ishimoto, M. (2005).Mapping of QTL associated with chilling tolerance during reproductivegrowth in soybean. Theor. Appl. Genet. 111, 851–861. doi: 10.1007/s00122-005-0007-2

Izawa, T., Oikawa, T., Sugiyama, N., Tanisaka, T., Yano, M., and Shimamoto, K.(2002). Phytochrome mediates the external light signal to repress FT orthologsin photoperiodic flowering of rice. Genes Dev. 16, 2006–2020. doi: 10.1101/gad.999202

Jia, H., Jiang, B., Wu, C., Lu, W., Hou, W., Sun, S., et al. (2014). Maturity groupclassification and maturity locus genotyping of early-maturing soybean varietiesfrom high-latitude cold regions. PLoS One 9:e94139. doi: 10.1371/journal.pone.0094139

Jiang, B., Nan, H., Gao, Y., Tang, L., Yue, Y., Lu, S., et al. (2014). Alleliccombinations of soybean maturity loci E1, E2, E3 and E4 result in diversityof maturity and adaptation to different latitudes. PLoS One 9:e106042. doi:10.1371/journal.pone.0106042

Kojima, S., Takahashi, Y., Kobayashi, Y., Monna, L., Sasaki, T., Araki, T., et al.(2002). Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transitionto flowering downstream of Hd1 under short-day conditions. Plant Cell Physiol.43, 1096–1105. doi: 10.1093/pcp/pcf156

Kong, F., Liu, B., Xia, Z., Sato, S., Kim, B. M., Watanabe, S., et al. (2010). Twocoordinately regulated homologs of FLOWERING LOCUS T are involved in thecontrol of photoperiodic flowering in soybean. Plant Physiol. 154, 1220–1231.doi: 10.1104/pp.110.160796

Kong, L., Lu, S., Wang, Y., Fang, C., Wang, F., Nan, H., et al. (2018). Quantitativetrait locus mapping of flowering time and maturity in soybean using next-generation sequencing-based analysis. Front. Plant Sci. 9:995. doi: 10.3389/fpls.2018.00995

Kurasch, A. K., Hahn, V., Leiser, W. L., Vollmann, J., Schori, A., Bétrix, C. A.,et al. (2017). Identification of mega-environments in Europe and effect of allelicvariation at maturity E loci on adaptation of European soybean. Plant CellEnviron. 40, 765–778. doi: 10.1111/pce.12896

Langewisch, T., Zhang, H., Vincent, R., Joshi, T., Xu, D., and Bilyeu, K. (2014).Major soybean maturity gene haplotypes revealed by SNPViz analysis of 72sequenced soybean genomes. PLoS One 9:e94150. doi: 10.1371/journal.pone.0094150

Li, J., Wang, X., Song, W., Huang, X., Zhou, J., Zeng, H., et al. (2017). Geneticvariation of maturity groups and four E genes in the Chinese soybeanmini core collection. PLoS One 12:e0172106. doi: 10.1371/journal.pone.0172106

Liu, B., and Abe, J. (2010). QTL mapping for photoperiod insensitivity of a Japanesesoybean landrace Sakamotowase. J. Hered. 101, 251–256. doi: 10.1093/jhered/esp113

Liu, B., Fujita, T., Yan, Z. H., Sakamoto, S., Xu, D., and Abe, J. (2007). QTLmapping of domestication-related traits in soybean (Glycine max). Ann. Bot.100, 1027–1038. doi: 10.1093/aob/mcm149

Liu, B., Kanazawa, A., Matsumura, H., Takahashi, R., Harada, K., and Abe, J. (2008).Genetic redundancy in soybean photoresponses associated with duplication ofthe phytochrome A gene. Genetics 180, 995–1007. doi: 10.1534/genetics.108.092742

Liu, W., Kim, M. Y., Kang, Y. J., Van, K., Lee, Y. H., Srinives, P., et al. (2011). QTLidentification of flowering time at three different latitudes reveals homeologousgenomic regions that control flowering in soybean. Theor. Appl. Genet. 123,545–553. doi: 10.1007/s00122-011-1606-8

Lu, S., Li, Y., Wang, J., Srinives, P., Nan, H., Cao, D., et al. (2015). QTL mappingfor flowering time in different latitude in soybean. Euphytica 206, 725–736.doi: 10.1007/s10681-015-1501-5

McBlain B. A., Hesketh J. D., and Bernard R. L. (1987). Genetic effects onreproductive phenology in soybean isolines differing in maturity genes. Can.J. Plant Sci. 67, 105–115. doi: 10.4141/cjps87-012

Nemoto, Y., Nonoue, Y., Yano, M., and Izawa, T. (2016). Hd1a, CONSTANSortholog in rice, functions as an Ehd1 repressor through interaction withmonocot-specific CCT-domain protein Ghd7. Plant J. 86, 221–223. doi: 10.1111/tpj.13168

Saindon, G., Voldeng, H. D., Beversdorf, W. D., and Buzzell, R. I. (1989). Geneticcontrol of long daylength response in soybean. Crop Sci. 29, 1436–1439.doi: 10.2135/cropsci1989.0011183X002900060021x

Samanfar, B., Molnar, S. J., Charette, M., Schoenrock, A., Dehne, F., Golshani, A.,et al. (2017). Mapping and identification of a potential candidate gene fora novel maturity locus, E10, in soybean. Theor. Appl. Genet. 130, 377–390.doi: 10.1007/s00122-016-2819-7

Schmutz, J., Cannon, S. B., Schlueter, J., Ma, J., Mitros, T., Nelson, W., et al.(2010). Genome sequence of the palaeopolyploid soybean. Nature 463, 178–183.doi: 10.1038/nature08670

Severin, A. J., Woody, J. L., Bolon, Y. T., Joseph, B., Diers, B. W., Farmer, A. D., et al.(2010). RNA-seq atlas of Glycine max: a guide to the soybean transcriptome.BMC Plant Biol. 10:160. doi: 10.1186/1471-2229-10-160

Song, Q., Jia, G., Zhu, Y., Grant, D., Nelson, R. T., Hwang, E. Y., et al. (2010).Abundance of SSR motifs and development of candidate polymorphic SSRmarkers (BARCSOYSSR_1.0) in soybean. Crop Sci. 50, 1950–1960. doi: 10.2135/cropsci2009.10.0607

Song, Q., Marek, L. F., Shoemaker, R. C., Lark, K. G., Concibido, V. C.,Delannay, X., et al. (2004). A new integrated genetic linkage map of the soybean.Theor. Appl. Genet. 109, 122–128. doi: 10.1007/s00122-004-1602-3

Song, Y., Ito, S., and Imaizumi, T. (2013). Flowering time regulation: photoperiod-and temperature-sensing in leaves. Trends Plant Sci. 18, 575–583. doi: 10.1016/j.tplants.2013.05.003

Takeshima, R., Hayashi, T., Zhu, J., Zhao, C., Xu, M., Yamaguchi, N., et al. (2016).A soybean quantitative trait locus that promotes flowering under long daysis identified as FT5a, a FLOWERING LOCUS T ortholog. J. Exp. Bot. 67,5247–5258. doi: 10.1093/jxb/erw283

Tamaki, S., Matsuo, S., Wong, H. L., Yokoi, S., and Shimamoto, K. (2007). Hd3aprotein is a mobile flowering signal in rice. Science 316, 1033–1036. doi: 10.1126/science.1141753

Tsubokura, Y., Matsumura, H., Xu, M., Liu, B., Nakshima, H., Anai, T., et al.(2013). Genetic variation in soybean at the maturity locus E4 is involved inadaptation to long days at high latitudes. Agronomy 3, 117–134. doi: 10.3390/agronomy3010117

Tsubokura, Y., Watanabe, S., Xia, Z., Kanamori, H., Yamagata, H., Kaga, A.,et al. (2014). Natural variation in the genes responsible for maturity lociE1, E2, E3 and E4 in soybean. Ann. Bot. 113, 429–441. doi: 10.1093/aob/mct269

Upadhyay, A. P., Ellis, R. H., Summerfield, R. J., Roberts, E. H., and Qi, A. (1994).Characterization of photothermal flowering responses in maturity isolines ofsoyabean [Glycine max (L.) Merrill] cv. Clark. Ann. Bot. 74, 87–96. doi: 10.1093/aob/74.1.87

Wang, D., Graef, G. L., Procopiuk, A. M., and Diers, B. W. (2004). Identification ofputative QTL that underlie yield in interspecific soybean backcross populations.Theor. Appl. Genet. 108, 458–467. doi: 10.1007/s00122-003-1449-z

Watanabe, S., Harada, K., and Abe, J. (2012). Genetic and molecular bases ofphotoperiod responses of flowering in soybean. Breed. Sci. 61, 531–543. doi:10.1270/jsbbs.61.531

Watanabe, S., Hideshima, R., Xia, Z., Tsubokura, Y., Sato, S., Nakamoto, Y.,et al. (2009). Map-based cloning of the gene associated with the soybeanmaturity locus E3. Genetics 182, 1251–1262. doi: 10.1534/genetics.108.098772

Watanabe, S., Tsukamoto, C., Oshita, T., Yamada, T., Anai, T., and Kaga, A. (2017).Identification of quantitative trait loci for flowering time by a combination ofrestriction site-associated DNA sequencing and bulked segregant analysis insoybean. Breed. Sci. 67, 277–285. doi: 10.1270/jsbbs.17013

Watanabe, S., Xia, Z., Hideshima, R., Tsubokura, Y., Sato, S., Yamanaka, N., et al.(2011). A map-based cloning strategy employing a residual heterozygous linereveals that the GIGANTEA gene is involved in soybean maturity and flowering.Genetics 188, 395–407. doi: 10.1534/genetics.110.125062

Frontiers in Plant Science | www.frontiersin.org 12 January 2019 | Volume 9 | Article 1867

fpls-09-01867 December 28, 2018 Time: 17:59 # 13

Zhu et al. E1Lb Represses Flowering Under LD

Wu, F., Price, B. W., Haider, W., Seufferheld, G., Nelson, R., and Hanzawa, Y.(2014). Functional and evolutionary characterization of the CONSTANS genefamily in short–day photoperiodic flowering in soybean. PLoS One 9:e85754.doi: 10.1371/journal.pone.0085754

Xia, Z., Watanabe, S., Yamada, T., Tsubokura, Y., Nakashima, H., Zhai, H., et al.(2012). Positional cloning and characterization reveal the molecular basis forsoybean maturity locus E1 that regulates photoperiodic flowering. Proc. Natl.Acad. Sci. U.S.A. 109, E2155–E2164. doi: 10.1073/pnas.1117982109

Xu, M., Xu, Z., Liu, B., Kong, F., Tsubokura, Y., Watanabe, S., et al. (2013).Genetic variation in four maturity genes affects photoperiod insensitivity andPHYA-regulated post-flowering responses of soybean. BMC Plant Biol. 13:91.doi: 10.1186/1471-2229-13-91

Xu, M., Yamagishi, N., Zhao, C., Takeshima, R., Kasai, M., Watanabe, S., et al.(2015). The soybean-specific maturity gene E1 family of floral repressorscontrols night-break responses through down-regulation of FLOWERINGLOCUS T orthologs. Plant Physiol. 168, 1735–1746. doi: 10.1104/pp.15.00763

Xue, W., Xing, Y., Weng, X., Zhao, Y., Tang, W., Wang, L., et al. (2008). Naturalvariation in Ghd7 is an important regulator of heading date and yield potentialin rice. Nat. Genet. 40, 761–767. doi: 10.1038/ng.143

Yamada, T., Hajika, M., Yamada, N., Hirata, K., Okabe, A., Oki, N., et al. (2012).Effects on flowering and seed yield of dominant alleles at maturity loci E2 andE3 in a Japanese cultivar, Enrei. Breed. Sci. 61, 653–660. doi: 10.1270/jsbbs.61.653

Yamanaka, N., Nagamura, Y., Tsubokura, Y., Yamamoto, K., Takahashi, R.,Kouchi, H., et al. (2000). Quantitative trait locus analysis of flowering time insoybean using a RFLP linkage map. Breed. Sci. 50, 109–115. doi: 10.1270/jsbbs.50.109

Yano, M., Katayose, Y., Ashikari, M., Yamanouchi, U., Monna, L., Fuse, T., et al.(2000). Hd1, a major photoperiod sensitivity quantitative trait locus in rice, isclosely related to the Arabidopsis flowering time gene CONSTANS. Plant Cell12, 2473–2483. doi: 10.1105/tpc.12.12.2473

Zhai, H., Lü, S., Wang, Y., Chen, X., Ren, H., Yang, J., et al. (2014). Allelic variationsat four major maturity E genes and transcriptional abundance of the E1 geneare associated with flowering time and maturity of soybean cultivars. PLoS One9:e97636. doi: 10.1371/journal.pone.0097636

Zhai, H., Lü, S., Wu, H., Zhang, Y., Zhang, X., and Yang, J. (2015). Diurnalexpression pattern, allelic variation, and association analysis reveal functionalfeatures of the E1 gene in control of photoperiodic flowering in soybean. PLoSOne 10:e0135909. doi: 10.1371/journal.pone.0135909

Zhang, W., Wang, Y., Luo, G., Zhang, J., He, C., Wu, X., et al. (2004). QTL mappingof ten agronomic traits on the soybean (Glycine max L. Merr.) genetic mapand their association with EST markers. Theor. Appl. Genet. 108, 1131–1139.doi: 10.1007/s00122-003-1527-2

Conflict of Interest Statement: The authors declare that the research wasconducted in the absence of any commercial or financial relationships that couldbe construed as a potential conflict of interest.

Copyright © 2019 Zhu, Takeshima, Harigai, Xu, Kong, Liu, Kanazawa, Yamadaand Abe. This is an open-access article distributed under the terms of the CreativeCommons Attribution License (CC BY). The use, distribution or reproduction inother forums is permitted, provided the original author(s) and the copyright owner(s)are credited and that the original publication in this journal is cited, in accordancewith accepted academic practice. No use, distribution or reproduction is permittedwhich does not comply with these terms.

Frontiers in Plant Science | www.frontiersin.org 13 January 2019 | Volume 9 | Article 1867